814 Proceedings of Royal Society of Edinburgh. [sess. 
icebergs, which in many cases are at a depth of 1500 to 1800 feet. 
The temperature of the mixture at the surface of contact falls ; the 
heat abstracted from the sea-water melts a portion of the iceberg, 
and a saline solution is produced, less salt, and therefore lighter 
than the water not in contact with the iceberg. This solution, 
being lighter, necessarily flows up along the sides of the iceberg to 
the surface, and its place is taken by new undiluted sea-water, which 
in its turn is cooled, diluted, and transferred to the surface, thus 
giving a practically constant temperature of 29 o, 0 in the neighbour- 
hood of the icebergs. The result is the production of an energetic 
engine of circulation and means of cooling and equalising the tem- 
perature of the water within the reach of icebergs. 
Another valuable research of Mr Buchanan’s deals with “ Simil- 
arities in the Physical Geography of the Great Oceans,”* in which he 
shows that the distribution of temperature from the surface of the 
ocean downwards follows the difference of density due to evapora- 
tion and dilution of the surface water. Wherever the density of the 
surface water of the sea is high, there the deeper are high tempera- 
tures found. On the other hand, where the density on the surface 
is low, as happens, for example, near rivers, there the high tempera- 
ture of the surface penetrates downward but a little way. Hence the 
important bearing of the surface waters of dry and rainy regions of 
the sea respectively on oceanic circulation. 
In the same paper are marked off, with a precision not previously 
attained, the low inshore temperature of the sea on the west of 
Africa from the Straits of Gibraltar to Cape Yerde, and again from 
about Loando to the mouth of the Orange River ; on the west of 
North America, from Vancouver Island to Cape St Lucas; and of 
South America, from the Gulf of Guayaquil to Chiloe. Prom 
Humboldt downwards these stretches of cold ■water have been gene- 
rally attributed to the flow equatorwards of cold oceanic currents 
from Antarctic and Arctic regions. Mr Buchanan shows, from the 
observations of navigators, the deep olive-green and often almost 
black appearance of the sea, and the abnormally low and practically 
uniformly distributed temperature of the sea through many degrees 
of latitude, that this cannot be the true explanation. He then 
points out that over all these extensive sea-boards the winds blow 
from the land westwards over the ocean, with the result that the 
cold water of greater depths is raised to the surface to take the 
place of the warmer waters of the surface, which are driven far to 
* Proc. Roy. Soc., vol. xxiii. p. 123, 1874, and Proc. Roy. Geog. Soc., Dec. 
1886. 
